[unreadable] [unreadable] This competing renewal application requests five years of support for the Conte Center for the Neuroscience of Mental Disorders (CCNMD): The Neurobiology of Suicidal Behavior. The Surgeon General notes, that 30,000 suicides per year in the United States and ten times that number of suicide attempts, requires a response that involves improving our knowledge regarding the relative importance of multiple predictors of suicide risk. The CCNMD employs a multidisciplinary approach to develop a predictive and explanatory model for suicidal behavior. Project 1: In the brainstem of depressed suicides we have found more serotonin neurons, more TPH, less serotonin transporter expression and 5-HT1A binding, all effects that enhance activity. The brainstem changes may be a response to less serotonin input and lower neurodensity in ventral prefrontal cortex of suicide and that may underlie increased impulsivity and suicide risk. We plan to map the circuitry of suicide by adding studies of anterior cingulate and amygdala, and to study neuron survival in prefrontal cortex. Project 2: CCNMD investigators have successfully tested a predictive stress-diathesis model where suicidal ideation is a proximal predictor and lifetime aggressivity/impulsivity is a trait predictor of suicide attempts in major depression. We now propose to evaluate potential causal factors (genes and adverse childhood experiences) and an intermediate biological phenotype, the HPA axis in a prospective study. Project 3: We will use PET to determine whether the postmortem receptor changes differentially associated with major depression and suicide can be distinguished in vivo. Effects of susceptibility gene expression on binding to 5-HTT, 5-HTiA and MAOA will be determined by PET and related to depression and suicidal behavior. Project 4 builds on our familial transmission results. This bottom-up familial transmission study will examine suicidality, aggression/impulsivity, serotonin susceptibility genes, and psychopathology in very high-risk families with depressed adolescent attempter probands. Projects 5 and 6 involve murine models of depression and aggression related to single genes and an inbred strain. Effects of altered gene expression at critical phases and of maternal deprivation on behavior, brain morphology, gene expression profiles, neurogenesis and biochemistry will be evaluated and compared with clinical findings. Project 7 proposed use of statistical model selection for suicide prediction and analysis of large postmortem and in vivo imaging data sets using statistical learning theory that allows the flexible Fitting of models with many predictors.